Spinal cord compression caused by metastatic cancer is the second most common neurologic complication of cancer and affects 5-15% of all cancer patients with solid tumors. With radiation therapy, the current standard treatment, less than half of patients are ambulatory after treatment. Uncontrolled studies have suggested that the combination of direct decompression of the spinal cord by surgical removal of the tumor actually causing the cord compression plus postoperative radiotherapy may be substantially better than radiation alone. However, only uncontrolled nonrandomized studies have been reported, and the apparent advantage of surgery found in these preliminary reports could be entirely due to preselection of good prognosis patients. We propose to do a multi-institutional prospective randomized trial to determine the actual value (if any) of direct decompressive surgery. Cancer patients with clinical and radiographic (or MRI) evidence of cord compression will be randomly assigned to treatment with either radiation alone (3,000 cGy) or direct decompressive surgery followed by radiotherapy (3,000 cGy). There will be a total of 200 patients (100 in each arm). The surgery will be performed using the most direct approach to the tumor, and the tumor causing the spinal cord compression will be removed. Intraoperative stabilization of the spine (using metallic rods or reconstruction of the vertebral body with acrylic) will be done if spinal instability is present. Major endpoints of the study will be ambulatory rate post treatment, pain relief, preservation of bowel and bladder function, recurrence rate, and time to recurrence. The main potential benefit of this study will be to determine the value of aggressive spinal surgery in the treatment of metastatic epidural cord compression. Given the poor results with radiation alone, any improvement in ambulatory rate or pain relief that results from surgery would be a major advance in the treatment of this fairly common oncologic problem. On the other hand, if aggressive spinal surgery is shown to be of no value, then the practice should be discontinued. Only a carefully controlled randomized trial will decide the issue.